From: Theo Markettos on
Nix <nix-razor-pit(a)esperi.org.uk> wrote:
> On 27 Feb 2010, Theo Markettos uttered the following:
> 344Mb caused by a couple of animations? I'd blame epiphany first. Gecko
> *loves* uploading crazy numbers of pixmaps to the server.

Ah, OK. It is a Gecko epiphany, rather than a Webkit one. I use Epiphany
as it's lighter than Firefox (the whole UI isn't written in Javascript for
Pete's sake).

> Bear in mind that X is rendering *everything* for *every* application: if
> e.g. you ask it to do 3D stuff and you don't have any hardware support,
> it'll have to render *that* in software as well, which is a real CPU pig.
> This is bad for all applications, not just X, because when X is using CPU
> it's not servicing requests from other apps.

I should probably mention that this is a laptop that gets regularly
suspended, so while the wallclock uptime may be 24 days it's actually
running for less than half that.

> Regarding X there are a lot of variables that can enormously affect your
> CPU consumption.
>
> What's the video card, what acceleration method is X using, and is the
> render extension accelerated?

On the laptop, nVidia G-Force FX. Using the nVidia binary driver,
I get about 600 fps in glxgears (with the CPU still loaded by Firefox) so I
think hardware acceleration is operational. Killing Firefox I get about
1200 fps.

The netbook is Intel 945 with whatever the native Ubuntu i945 support is.
When quiet (browsers killed) I get 140 fps (with part of the window obscured
as UNR maximises it).

How do I find out the acceleration method? My xorg.conf is fairly spartan.

glxinfo says, for the laptop:
OpenGL vendor string: NVIDIA Corporation
OpenGL renderer string: GeForce FX Go5200 32M/64M/AGP/SSE2
OpenGL version string: 2.1.2 NVIDIA 173.14.16
OpenGL shading language version string: 1.20 NVIDIA via Cg compiler

for the netbook:
OpenGL vendor string: Tungsten Graphics, Inc
OpenGL renderer string: Mesa DRI Intel(R) 945GME GEM 20090326 2009Q1 RC2
x86/MMX/SSE2
OpenGL version string: 1.4 Mesa 7.4

Is that software rendering? Or does Mesa have a hardware driver for i945?

I tried adding:
Driver “intel”
Option “AccelMethod” “UXA”
VideoRam 261376

to xorg.conf, but then I got 18 fps from glxgears maximised on the netbook.
(All times here are with it on battery). glxinfo still says I'm using Mesa.

I should add that Flash video (YouTube, iPlayer etc) does struggle, so there
may well be something up with the video acceleration.

> It gets slow because, though it is unable to swap out dirty data, it
> *can* throw out clean file-backed pages (largely text pages from
> binaries and shared libraries). Unfortunately because it *can't* swap
> out dirty data, it has to leave it all in memory, even the hundreds of
> kB of gettext-related rubbish mapped into virtually every app, and the
> relocations for every single app you're running, even if they're never
> accessed: and in their place it may have to throw out very popular pages
> from binaries, only to have to read them back in again a fraction of a
> second later.

Ah, right. So more RAM might help.

> Even with overcommit on you sometimes might get spontaneously killed for
> being out of memory. The fork()/exec() case will kill the child with no
> way to stop it, and if you run out of memory expanding the stack (which
> is always lazily allocated), you're as dead with overcommit off as with
> it on. As such, I consider overcommit-off to bring a false sense of
> security at huge cost.

Thanks for the background - I hadn't understood the options. Looks like
overcommit is the sort of thing you might do on a server. Not sure it's
appropriate for a laptop.

Theo
From: Theo Markettos on
Gordon Henderson <gordon+usenet(a)drogon.net> wrote:
> In article <DAy*WFM4s(a)news.chiark.greenend.org.uk>,
> Theo Markettos <theom+news(a)chiark.greenend.org.uk> wrote:
>
> >That's why I was wondering about adding RAM. But folks seemed to suggest
> >it was disc that's the bottleneck, which is why I'm confused.
>
> Probably mostly me i(suggesting disk) - as that's what I see on my own
> (and wifeys) AAO netbooks with SSD - however it's now become aparently
> from other posts that you actually "use" this box! As in use it for day
> to day use rather than just for quick little jobs on customer sites -
> which is primarly what mine is for - boot it up, run firefox and a fiew
> xterms/ssh sessions and that's mostly it.

It's not that different - only I tend not to turn it off for weeks at a
time, and run dozens of browser tabs. Which seems to slow it down rather
too much.

> It's almost as if it's running in old-fashioned PIO mode...
>
> However, even with more RAM, I still think it'll go slow - on my AAO it
> stalls totally when it writes to it's SSD - and I have an access LED to
> prove it! (I'm somewhat surprised the Dells didn't provide this)

Yes, it's rather annoying. Particularly since there are two power LEDs.
There's a hack to convert one to an activity light, but I have more pressing
things on my taking-apart-hardware queue.

> I can load a page in firefox and it stalls half way through - disk LED
> comes on - then it loads, then stalls again - disk LED on again and so
> on. My thoughts are that firefox is writing cache, history & cookies to
> disk and doing an fflush to make sure there're there so when it crashes
> it has more on-disk to pick-up from where it left off.

Interesting. I wonder why it's so paranoid about crashing? ;-)
I do rely on recovering old tabs after crashes, though, so I'd miss this.

> Once a page is laoded it's usually remarkably quick - however I have a
> custom compiled kernel and took care to remove as much unused userland
> utilities, etc. as I could.

That sounds like rather more pain than I have time for...

> A quick & dirty test:
>
> On the AAO:
>
> # time dd if=/dev/zero of=testfile bs=1M count=1000
> 1000+0 records in
> 1000+0 records out
> 1048576000 bytes (1.0 GB) copied, 104.777 s, 10.0 MB/s
>
> real 1m44.817s
> user 0m0.008s
> sys 0m7.104s

I've done it for 500 blocks (as I don't have enough space otherwise):

524288000 bytes (524 MB) copied, 42.078 s, 12.5 MB/s
real 0m42.090s
yser 0m0.012s
sys 0m5.796s

Looks slightly faster.

> Raw disk speed doesn't tell us everything though - 35MB/sec on my desktop
> is slow by todays standards - the 1990's are calling for that SSD... Doing
> an fflush is probably going to cause a lot of small writes to take place -
> probably the worst thing for an SSD top have to cope with.

Not sure... there are no seek times on SSD which I'd have thought would slow
down small writes (though some flash isn't as fast at seeking as you'd
expect)

> When searching for ways to speed up the AAO, I came across several
> articles complaining about the slow SSDs - and it turns out Acer used
> 2 brands - a cheaper and slower one in the early days and a slightly
> faster one in later days... There is even an article about replacing
> the SSD with a much faster CF card!

Yes, that's the same with the Dell - 'RunCore' replacement SSDs are a lot
faster. But cost about half the price of the netbook again to buy
aftermarket. I did wonder about trying to fit internal USB flash, but there
isn't much space for a hub chip and the flash (most hacks use the empty WWAN
slot, which is occupied on mine).

> So.. Not sure where I'd go from here - if this was a day to day workhorse,
> personyll, I'd probably get something with a much faster disk - or look
> in installing a faster SSD or CF type of thing - or even move to a spinny
> disk, if possible...

Looks like the video driver might be the culprit... will investigate
further.

Theo
From: Darren Salt on
I demand that Theo Markettos may or may not have written...

> Nix <nix-razor-pit(a)esperi.org.uk> wrote:
[snip]
>> What's the video card, what acceleration method is X using, and is the
>> render extension accelerated?

> On the laptop, nVidia G-Force FX. Using the nVidia binary driver, I get
> about 600 fps in glxgears (with the CPU still loaded by Firefox) so I think
> hardware acceleration is operational. Killing Firefox I get about 1200
> fps.

> The netbook is Intel 945 with whatever the native Ubuntu i945 support is.
> When quiet (browsers killed) I get 140 fps (with part of the window
> obscured as UNR maximises it).

> How do I find out the acceleration method? My xorg.conf is fairly spartan.

Looks like you've found out some of it (below). For the rest, check
/var/log/xorg.0.log.

> glxinfo says, for the laptop:
> OpenGL vendor string: NVIDIA Corporation
> OpenGL renderer string: GeForce FX Go5200 32M/64M/AGP/SSE2
> OpenGL version string: 2.1.2 NVIDIA 173.14.16
> OpenGL shading language version string: 1.20 NVIDIA via Cg compiler

No idea about that one.

> for the netbook:
> OpenGL vendor string: Tungsten Graphics, Inc
> OpenGL renderer string: Mesa DRI Intel(R) 945GME GEM 20090326 2009Q1 RC2
> x86/MMX/SSE2
> OpenGL version string: 1.4 Mesa 7.4

> Is that software rendering? Or does Mesa have a hardware driver for i945?

Hardware. Not KMS, though, IIRC.

[snip]
> I should add that Flash video (YouTube, iPlayer etc) does struggle, so
> there may well be something up with the video acceleration.

Lack of support for hardware decoding, or (given that those aren't MPEG2) no
hardware decoding. Still, if it's using Xv for display, you'll get some
acceleration there.

[snip]
--
| Darren Salt | linux at youmustbejoking | nr. Ashington, | Doon
| using Debian GNU/Linux | or ds ,demon,co,uk | Northumberland | Army
| + http://www.xine-project.org/

You are taking yourself too seriously.
From: Tony Houghton on
In <50F72FAE84%news(a)youmustbejoking.demon.cu.invalid>,
Darren Salt <news(a)youmustbejoking.demon.cu.invalid> wrote:

>> I should add that Flash video (YouTube, iPlayer etc) does struggle, so
>> there may well be something up with the video acceleration.
>
> Lack of support for hardware decoding, or (given that those aren't MPEG2) no
> hardware decoding. Still, if it's using Xv for display, you'll get some
> acceleration there.

Flash definitely doesn't support hardware scaling. And judging by its
performance its software scaler appears to be written in a particularly
poor dialect of BASIC running on an interpreter implemented in Flash's
AVM.

--
TH * http://www.realh.co.uk
From: alexd on
On Mon, 01 Mar 2010 14:08:55 +0000, Theo Markettos wrote:

> I use Epiphany as it's lighter than Firefox (the whole UI isn't written
> in Javascript for Pete's sake).

As Firefox's job is to take Javascript [and XML] as input and render it
as output, what's your objection here?

--
<http://ale.cx/> (AIM:troffasky) (UnSoEsNpEaTm(a)ale.cx)
20:33:28 up 16 days, 1:24, 4 users, load average: 0.12, 0.06, 0.01
DIMENSION-CONTROLLING FORT DOH HAS NOW BEEN DEMOLISHED,
AND TIME STARTED FLOWING REVERSELY